Prompt Obfuscation And Evasion Technique Catalog

via “obfuscation detection and deobfuscation assistance”

Show HN: Ghidra MCP Server – 110 tools for AI-assisted reverse engineering

Unique: Combines pattern detection with heuristic analysis to identify obfuscation techniques and provide deobfuscation guidance, rather than just flagging suspicious code

vs others: Provides actionable deobfuscation hints alongside detection, enabling LLMs to assist in understanding obfuscated code

via “prompt-obfuscation-and-evasion-technique-catalog”

LEAKED SYSTEM PROMPTS FOR CHATGPT, CLAUDE, GEMINI, GROK, PERPLEXITY, CURSOR, LOVABLE, REPLIT, AND MORE! - AI SYSTEMS TRANSPARENCY FOR ALL! 👐

Unique: Documents obfuscation techniques (leetspeak, special characters, context manipulation) as reproducible attack patterns with model-specific effectiveness data, rather than treating them as one-off exploits. The repository tracks which obfuscation strategies work against which models and versions.

vs others: Provides a curated, model-specific catalog of obfuscation techniques with effectiveness metrics, whereas most security research on prompt injection evasion is scattered across informal disclosures without systematic evaluation.

via “security tool evasion and detection avoidance guidance”

MCP server: pentest-copilot

Unique: Provides LLM-driven evasion guidance based on identified security tools, allowing Claude to recommend context-aware evasion strategies rather than generic techniques

vs others: Tailors evasion recommendations to specific target security posture compared to generic evasion guides, with LLM-driven analysis of tool-specific detection mechanisms